Illustration by Jørgen Stamp digitalbevaring.dk CC BY 2.5 Denmark
Overview
This case study provides a brief novice to intermediate level overview summarised from the DPC Technology Watch Report on Preserving Moving Picture and Sound. Five "mini case studies" of UK collections that have run preservation and access projects for sound and moving image content are included. The report itself provides a "deep dive" discussing a wider range of issues and practice in greater depth with extensive further reading and advice (Wright, 2012). It is recommended to readers who need a more advanced level briefing on the topic and practice.
Introduction
The audiovisual domain is unique in that digitization is routinely critical to preservation. Audiovisual digitization for preservation is so pervasive that the two words have come to be used interchangeably. Audio and video need digitization for the very survival of their content, owing to the obsolescence of playback equipment and decay and damage of physical items, whether analogue or digital. The basic technology issue for collections of moving images and sound is the necessity to digitize all content currently sitting on shelves. Film on shelves can be conserved (unless it is already deteriorating), but still needs digitization to provide access.
A vital issue in preservation is access: motivation and funding for digitization purely for preservation purposes is difficult, if not impossible. There is great public, institutional and educational interest in the audiovisual record of the twentieth century. Creating access to that record is the key to obtaining the support needed for the digitization and preservation of the content.
The landscape for 'moving pictures and sound' is complicated: physically, there are large differences between audio, video and film recordings. The formats and record/playback equipment are completely separate; the digitization procedures are different; the digital files have different wrapper formats and metadata (with some overlaps); and the storage requirements differ, with video taking roughly 100 times as much storage per second of material as does audio, and high resolution digital film taking roughly 10 times more storage than video.
In addition culturally and economically, there are significant preservation and curation differences between collections from:
- commercial media industries – music, cinema and commercial broadcasting where preservation needs a commercial justification, a business case;
- public bodies – public service broadcasting, academic collections and heritage institutions such as national museums, libraries and film institutes where preservation needs a cultural heritage justification, though increasingly this sector also needs a business case;
- technical areas such as medicine, geology and surveillance, where recordings of images or of seismic events are raw data, kept as medical records or for reprocessing; and
- other – a wide range of independent collections, ranging from individual efforts to material gathered by non-profit specialist institutions (for example, steam engine clubs or ethnological research) that do not fall into any of the above categories, though their material may eventually end up being donated to a public collection.
Within the landscape is a range of technologies including engineering, computing, Internet technology, archiving, media management, museum collections management, curation, preservation, access, knowledge management and resource discovery.
Technical challenges
Audiovisual recordings are surrogate reality. The technology allows the listener and viewer to get a sensation of what a situation sounded and looked like, but the technology actually only captures the sequence of light patterns or sound pressures acting on the recording instrument (camera, microphone). These patterns (for film) and signals (for video and audio) are more like data than like artefacts. The preservation requirement is not to keep the original recording media, but to keep the data, the information, recovered from that media.
A key technology issue is moving digital content from carriers (such as CD and DVD, digital videotape, DAT and minidisc) into files. This digital to digital 'ripping' of content is an area of digital preservation unique to the audiovisual world, and has unsolved problems of control of errors in the ripping and transfer process.
The final technology area is digital preservation of the content within the files that result from digitization or ripping, and the files that are born digital. While much of this preservation has problems and solutions in common with other content, there is a specific problem of preserving the quality of the digitized signal that is again unique to audiovisual content. Managing quality through cycles of lossy encoding, decoding and reformatting is one major digital preservation challenge for audiovisual files. The other issue is managing embedded metadata.
For three decades for audio, and for at least two decades for video, archives have been digitizing their analogue content for preservation and access. The problem areas are:
- successful playback of the originals, in order to get an optimal signal to digitize;
- standards: what compression level, encoding method and file format to use; and
- efficiency: digitizing the existing analogue materials fast enough and economically enough to cope with the size and urgency of the problem.
Stages in sound and moving image digital preservation
For sound and moving image preservation, the following stages in the overall process need to be kept clear:
- signal: the audio from a microphone, the video signal coming out of a video camera. These signals have physical properties (bandwidth; dynamic range) that can be defined and measured. The quality of a recording and the success or failure of any process of copying, digitization or preservation can be reduced (in large part) to how well that process maintains these two physical properties of the original signal;
- recording of a signal onto a carrier (also called support, physical medium or recording format). For a century, the methods of capturing a signal were tied to the carrier of the signal: a wax cylinder, film reel or videotape. Digital technology produces recordings that are independent of carriers. Carrier independence is liberation: discs, tapes and films deteriorate or get damaged. Born digital recordings are liberated from these carrier-based problems, leading to a desire to liberate analogue recordings by digitization;
- digitization: analogue recordings can be played back and recorded onto a new carrier, or digitized and so released from carrier dependence. Digitization has to ensure that the digital version has the same bandwidth and dynamic range as the original, to capture the original quality; and
- digital preservation of the digital representation of a signal, meaning preserving the numbers, but also preserving the technology needed to decode (render) the numbers. Audiovisual content has a particular problem. The coding of the signal can be a compromise, not actually capturing the full signal, but instead losing some of it (lossy encoding) to get a more compact representation, thus reducing storage and transmission costs. Unfortunately coders/decoders (codecs) go out of use, and are replaced by newer technology. The file format holding the coded signal, the wrapper, is also subject to obsolescence. The failure and obsolescence of storage technology and the obsolescence of encode/decode methods and wrapper formats are major digital preservation problems for audiovisual content.
Access and rights
Sound and moving picture content arising from cinema, broadcasting and the commercial music industry is constrained by rights issues. Music has copyright protection for the composer and for the physical object containing a performance (so-called magnetic copyright). Cinema productions are protected, and music used in a film retains its separate protections. Broadcasting is even more complicated, as all the parties involved in a production may have rights in future exploitation subsequent to the one or two transmissions that were specified in typical contracts. These rights are seen as protection by rights holders, but are also seen as restrictions on access. The situation for a public broadcaster is particularly difficult. The public invariably feel that any production by a public broadcaster has already been paid for by them, is already publicly owned and should be available for public access. Unfortunately that understandable feeling is not the same as the legal definition governing when a work enters the public domain (usually determined by expiry dates on copyright and other rights).
Case study 1: The Open University (OU) Access to video assets project
This is an access and re-use project. The focus is to digitize (where necessary) audiovisual assets previously created by the OU, and place them in an asset management system so that current OU teaching and other activity can find and use these assets. Preservation is a by-product of the project rather than an end in itself. This project provides an important example of combining preservation of content with use of content, something of value to the institution in order to obtain a budget and deliver a benefit. The project was presented at the DPC Briefing Day 'Preserving Digital Sound and Vision'. The project digitized 1,200 videotapes and films, and placed the results in a Fedora digital repository. Also, 145,000 pages of documentation were digitized, providing the overall educational framework around the 1,200 items, giving them context and enhancing their ability to be re-used. The user interface provides granularity and time-based navigation. Overall this project is an outstanding example of best practice.
Case study 2: British Library Archival sound recordings project
This is a JISC-supported preservation and educational access project that ran (in its initial phase) from 2004 to 2006. A second phase added further material. Nearly 50,000 recordings of speech, music and sounds of 'human and natural environments' were digitized and placed online. The online catalogue is open to all and licensed UK further or higher education institutions can also listen to the audio. Anyone can listen to 2,000 of the items (or any of them by attending the British Library reading room in London). The differences in access between educational institutions and the general public reflects the overall issue of rights as the one remaining constraint on open access to audiovisual materials in public institutions.
Case Study 3: Imperial War Museum PSRE project
The Imperial War Museum has one of the UK's major film collections. It has been collecting film since its founding in 1919, beginning with footage from the Great War that led to the institution's founding. The Public Sector Research Exploitation (PSRE) fund made an award of nearly £1 million for cataloguing, digitization and online access (to the catalogue and the footage). The project ran from 2006 to 2009 and is of particular interest in that it is specifically aimed at commercial exploitation of a collection, and at sustainable business models around digitization and web access. The result is a website (http://film.iwmcollections.org.uk/) where anyone can view content in low quality; pull documents, stills and key frames into a lightbox; and fill a shopping basket to then purchase content.
Case Study 4: British University Film and Video Council Newsfilm Online project
This is another project with JISC sponsorship. For four decades to 1960 newsreels shown in cinemas were the main way for the general public to see moving images of current events. The initial project ran from 2004 to 2008. The results are available through a website which, as for the BL Archival Sound Recordings project, has full functionality for registered universities and colleges. The general public can see the full catalogue and can see a single key frame for each item. Since the original phase of the project, the content has been augmented by ITN/Reuters news covering the events from decades after the decline of newsreels. Newsreel items are short: the initial project provided 3,000 hours of content, but that represented 60,000 items. In addition, as with the Open University project, documentation was also placed online for context and to support search and retrieval: 450,000 pages of bulletin scripts.
Case Study 5: BFI and Regional Film Archives Screen Heritage UK (SHUK) project
SHUK is a large (£22.8 million) and complex project (involving 12 regional film archives in addition to the BFI). The project was complicated by changes in the structure and funding of the BFI, as well as a change of government and a raft of other issues. Nevertheless the project has produced major achievements:
- conservation, not digitization: construction of a £6-million vault for film conservation;
- digitization: film scanning and digital storage equipment for the regional film archives;
- access: online catalogues of regional film archive content, available to the general public.
SHUK launched on 5 September 2011 with a BBC BFI joint production, The Reel History of Britain (SHUK, 2011).
Conclusions
The basic technology issue for collections of moving images and sound is the necessity for digitization of all content that is currently sitting on shelves. Audio and video need digitization for their very survival, owing to obsolescence and decay of physical items, whether analogue or digital. Film on shelves can be conserved (unless it is already deteriorating) but needs digitization for access.
Playback for preservation-quality digitization implies the need for optimal recovery of the original quality, which requires professional equipment and experience. The major technical obstacle is that, for many physical formats, the needed equipment is largely obsolete, meaning that parts and repairs and skilled operators are in increasingly short supply. The urgent recommendation is, do not wait! Audiovisual holdings need to be documented and made part of a preservation plan.
The situation for sound heritage is clear. The digitization standards, encoding, wrapper and metadata are all agreed and well documented in IASA TC-04 (IASA, 2009). Uncompressed audio in the Broadcast Wave Format (BWF) wrapper is widely used and well supported. There is no reason for the basic encoding to ever be changed, though the BWF wrapper may eventually become obsolete. The only significant problem is the failure of some standard audio applications to handle embedded BWF metadata correctly (ARSC, 2011). All archives need to be aware of the risk of loss of embedded metadata. The situation for video is complex, but there is a PrestoSpace roadmap for guiding choices on the digitization of various legacy formats. There is advice from the PrestoCentre and from JISC Digital Media on the digital preservation of the resultant files. A big challenge is a registry of applications that work properly on embedded video metadata, where the diversity is huge. There is no single agreed wrapper, metadata standard or even encoding standard, and the change from standard definition to high definition brings a new set of applications, wrappers and encodings.
There is emerging technology that can improve audio (capture of the bias tone and consequent removal of temporal variation) and video transfers (direct digitization of the RF signal from the read head), which could be useful in those cases where current technology fails. So the recommendation is not to wait until such technology is further advanced and more widely available. If there are playback problems that cannot be resolved, the original audio or video format should be kept so that such advanced technology can be applied in the future.
Quality checking of the results of digitization remains an issue for video. There is a need for effective integration of signal processing technology with human checking in order to produce a really efficient method of quality control within a preservation factory approach. Quality checking is equally relevant to digital preservation – any changes or migrations due to digital obsolescence need to be checked for preservation of signal quality. Again, a purely manual approach does not scale (to the tens of millions of hours of audiovisual content in European collections), while purely algorithmic substitutes for 'looking and listening' have never been completely successful and remain an area where further research is needed.
Resources
Wright, R., 2012. Preserving Moving Pictures and Sound DPC Technology Watch Report 12-01 March 2012
http://dx.doi.org/10.7207/twr12-01
This report is for anyone with responsibility for collections of sound or moving image content and an interest in preservation of that content. New content is born digital, analogue audio and video need digitization to survive and film requires digitization for access. Consequently, digital preservation will be relevant over time to all these areas. The report concentrates on digitization, encoding, file formats and wrappers, use of compression, obsolescence and what to do about the particular digital preservation problems of sound and moving images (33 pages).
SHUK, 2011. Screen Heritage UK Marks new Era for Britain's Film Archives
BFI Press release. 8 pages
IASA 2009 IASA TC-04, Guidelines on the Production and Preservation of Digital Audio Objects (IASA-TC 04 Second edition 2009) Canberra, IASA.
http://www.iasa-web.org/audio-preservation-tc04
This is the standard guide to digitization of audio, and the sections on metadata and digital storage are of value to all forms of digital media.
Casey, M. and Gordon, B., 2007. Best Practices for Audio Preservation. Bloomington, Indiana University Bloomington.
http://www.dlib.indiana.edu/projects/sounddirections/papersPresent/
Another audio resource (that also includes a range of digitization software tools) comes from the Sound Directions project of Harvard and Indiana Universities: much is also relevant to video digitization. (160 pages)
Digital Preservation Coalition Briefing day on Preserving Digital Sound and Vision, April 2011
https://www.dpconline.org/events/past-events/preserving-digital-sound-and-vision-a-briefing
This DPC briefing day in April 2011 provided a forum to review and debate the latest development in the preservation of digital sound and vision. Seven presentations (including the Open University) are linked from the programme and available to download.
ARSC Technical Committee, 2011. Study of Embedded Metadata Support in Audio Recording Software. Association of Recorded Sound Collections.
http://www.arsc-audio.org/pdf/ARSC_TC_MD_Study.pdf
A study of support for embedded metadata within and across a variety of audio recording software applications. The findings raise serious concerns, particularly for the archiving and preservation communities who rely on embedded metadata for interpretation and management of digital files representing preserved content into the future. (21 pages)
AVPreserve
US based media and information management consulting firm. It website provides a range of resources for AV preservation.
BUFVC NewsFilm online Project
British Film Institute
the British Film Institute can advise on film and also on video – they hold a lot of video, and have a Curator for Television. Its remit is collection and preservation of film and television, and technical advice.
British Library Sound Archive
https://www.bl.uk/subjects/sound#
General technical advice on audio preservation is available from the British Library Sound Archive. Its remit is collection and preservation of all forms of audio, and technical advice.
Film Archives UK
Collection and preservation of general audiovisual content of regional significance in the UK
JISC Digital Media
Advice and training on still images, moving images and sound. This includes their InfoKits for Digital File Formats, Digitisation funding and sustainability, and High Level Digitisation Guide for Audiovisual Resources.
PrestoCentre
Website provides audiovisual information, resources and advice. Access has recently been extended so that all resources are now freely available to all.
Sustaining Consistent Video Presentation
http://www.tate.org.uk/research/publications/sustaining-consistent-video-presentation
This technical paper addresses approaches to identifying and mitigating risks associated with sustaining the consistent presentation of digital video files. Originating from two multi-partnered research projects – Pericles and Presto4U – the paper was commissioned by Tate Research and is intended for those who are actively engaged with the preservation of digital video.
JISC 2009 - Archival Sound Recordings Showreel
https://www.youtube.com/watch?v=KPy9ZqWEHog
Engaging short video on British Library archival sound recordings project published on 22 Jun 2009. (6 mins 11 secs).
Further case studies
Podcasts in the Archives: Archiving Podcasting Content at the University of Michigan
http://files.archivists.org/pubs/CampusCaseStudies/CASE12.pdf
In this Society of American Archivists campus case study Alexis. A. Antracoli, University of Michigan, examines the challenges involved in developing best practices and workflows for archiving and preserving podcasting content. One major issue involved establishing standards of practice for ingest, storage, and access, especially the generation and storage of appropriate descriptive, technical, and preservation metadata. Another challenge centered around developing the necessary technological infrastructure to support an Open Archives Information System (OAIS)-compliant system. 2010. (14 pages).
References
ARSC Technical Committee, 2011. Study of Embedded Metadata Support in Audio Recording Software. Association of Recorded Sound Collections. Available: http://www.arsc-audio.org/pdf/ARSC_TC_MD_Study.pdf
IASA, 2009. IASA TC-04, Guidelines on the Production and Preservation of Digital Audio Objects, IASA-TC 04 Second edition 2009, Canberra, IASA. Available: http://www.iasa-web.org/audio-preservation-tc04
SHUK, 2011. Screen Heritage UK Marks new Era for Britain's Film Archives. Available: http://www.bfi.org.uk/sites/bfi.org.uk/files/downloads/bfi-press-release-screen-heritage-uk-marks-a-new-era-for-britains-film-archives-2011-09-01.pdf
Wright, R., 2012. Preserving Moving Pictures and Sound DPC Technology Watch Report 12-01 March 2012. Available: http://dx.doi.org/10.7207/twr12-02